This invention relates generally to clinical information systems, and more directly to clinical decision support systems.
In a clinical decision support system, a “rule” is intended to indicate what procedure or treatment to perform for a particular patient being monitored. For example, if a patient's measured potassium level dropped below 5, then a physician or nurse would be notified with this information. Thus, a rule would be written specifying that, when a patient's potassium level is below 5, send an email or a page to a particular nurse.
Rules written in decision support systems are typically written in a computer language such as JAVA®. Rules can depend upon more than one data input. If all of the data inputs are not available, the decision support system cannot process the rule accurately. This causes the decision support system to skip the rule or flag an error for an administrator to look at, at some later time. However, if all of the data inputs had been present, an alert might have been sent which might have notified clinicians of a potential problem.
For example, if a rule was set to watch both potassium and sodium levels in a patient and to signal an imbalance if the sodium level goes up and the potassium level goes down, there may be nothing to indicate an error if there are no measurements for sodium input to the decision support system. Also for example, if a new potassium level comes into the database and the sodium level measurement was from the previous day rather than within the last half hour as desired, the decision support system may not provide any decision or notify anyone immediately, as known decision support systems are configured to inform caregivers when certain events are happening. If a decision cannot be made, to prevent false notifications, no notifications are sent out, even though a problem may exist.
Although it may be possible to provide popup data windows on decision support systems, such windows are invasive and are unlikely to provide a suitable workflow for clinicians.